Traffic management

Table of Contents

Commentary 2 Performance of Advanced Traffic Signal Control Strategies

Gartner et al. (1995) investigated the development of advanced traffic signal control strategies and considered the performance realised by their implementation. To compare different control strategies, a hierarchy was developed based on different generations of traffic signal control. This ranged from simpler control strategies, such as fixed-time control (i.e. 0-GC) to more complex strategies, such as fully adaptive traffic control (i.e. 3-GC).

A literature review was conducted of published reports noting the benefits realised when implementing more complex traffic signal control strategies. The results of the comparison are illustrated in Figure C2 1 with:

As shown in Figure C2 1(b), a number of field implementations of more complex control strategies reported did not lead to improved performance. The authors identified a number of factors that may have contributed to the degradations in performance such as:

  • inaccurate prediction of cycle time resulting in strategies that were unable to respond to rapid changes in traffic flow
  • frequent transition of signal timing resulting in considerable delays
  • local optimisation of cycle times resulting in dis-benefits due to a loss of synchronisation among intersections in a network.

The authors stressed the importance of continuous monitoring and evaluation of control strategies to ensure that optimal performance is realised when applying more complex control strategies.

Whilst this example is specific to traffic signal control systems, it presents similar issues that need to be considered for other forms of ITS. Namely that substantial effort may be required to configure systems to operate optimally. Additionally, regular review of configuration data may often be required to ensure a system continues to operate optimally in the face of changing traffic environments.

Figure C2 1: Expected and reported performance for traffic signal control generations

Source: Gartner et al. (1995).

[Back to body text]